In a cloud environment, control loop automation can be categorized into open loop or closed loop systems. Open loop systems capture telemetry and diagnostics information from the underlying cloud infrastructure (e.g. syslog, SNMP, fault and performance management events), perform a set of analytics and provide reporting or alarms to the operations team. Closed loop systems continuously monitor the system for fault, performance, security, etc. related problems and compute a set of signatures based on the detected anomalous condition. These signatures are then interpreted and appropriate corrective actions are recommended to repair the system. Once the system has been repaired, a monitoring application checks the status to see if the system responded to alleviate the detected problem.
Control loop systems play a vital role in processes implemented in a cloud based environment. For example, closed and open loops have been implemented to support fault management for limited virtual network functions such as a wide-area virtual network management framework which acts as an infrastructure provider to manage the physical resources and service provider to build and manage virtual networks on the physical resources.
Control loops in cloud based environments have been developed as independent applications. However, traditional monolithic architectures for control loop are not flexible enough to meet the demands of the operators of the future. For example, control loops are traditionally written with most of the critical logic embedded in one software module.
There is a need to develop a platform that will enable faster development of scalable control loop systems.